Background/Purpose
Studies have proved the critical role of environment in determining people's physical activities (PA) participation; but commonly used PA measures can only measure how active a person is without providing information about where and when a specific activity occurs. Zhu (2003) proposed tracking of people's physical activity space (PAS) to better understanding of the interaction between their PA and the environment. A combination of GPS and accelerometer was used to track PAS, but the cost was high and the data management was not easy. Cell phones are almost ubiquitous these days, e.g., more than 66% of the U.S. population owns a cell phone, that is around 200 million users (Katz, 2005). Furthermore, a GPS-enabled cell phone could provide information about the user's location, which provides new potential for the study of PAS in a high-accuracy, low–cost and convenient way. The purposes of this study were to examine GPS capability of modern cell phones in gathering data about people's PAS.
Method
We selected the new Google phone (G1) for the case study because of several features: (a) The simplicity of its new operating system (OS) in developing new applications, (b) Existence of free GPS tracking applications, and (c) low cost compared to phones with similar functions. Furthermore, its selected GPS tracking application (InstaMapper) could provide real-time location data every 5 seconds even when the screen is off. Finally, it could provide up to 5 hours of data streaming with a single battery charge. The phone's capabilities were tested for over a month in various conditions and a number of features related to PAS were examined, including phones capabilities in recording heading direction and speed. In addition, we were able to add a filtering criterion to delete recorded GPS distorted areas and increase the accuracy to 5 meters.
Analysis/Results
The G1 phone was able to record data continuously during the study period. Recorded tracking data were accessible in real-time over the web and the provided Google Maps API (application programming interface) made it easy to work with the data. Using developed filtering algorithm further enhanced the results and enabled the detection of no movement, walking, running, and driving.
Conclusions
G1 phone can track a person's location in an easy and practical way. By integrating a PA measure, tracking of PAS should be possible. However, the G1 phone is unable to track location indoors.